International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 10 | Oct 2025
p-ISSN: 2395-0072
www.irjet.net
Automated Detection of Thyroid Using Convolutional Neural Networks - A Survey Tanmay Bhale1, Prof. Pramila M Chawan2 1MTech Student, Dept. of Computer Engineering & IT, VJTI College, Mumbai, Maharashtra, India 2Associate Professor, Dept. of Computer Engineering & IT, VJTI College, Mumbai, Maharashtra, India
---------------------------------------------------------------------***--------------------------------------------------------------------from medical images, enabling them to perform automated Abstract - Accurate and early diagnosis of thyroid nodules is diagnosis of thyroid nodules
critical for effective management of thyroid diseases. Traditional diagnosis relies heavily on expert interpretation of ultrasound images, often resulting in subjectivity and variability. With advances in artificial intelligence, convolutional neural networks (CNNs) have emerged as powerful tools for automated image analysis and disease recognition. This paper surveys the recent developments in using CNNs for thyroid nodule detection, classification, and risk stratification, highlighting their performance compared to traditional methods and expert radiologists. We also discuss challenges in model generalization, dataset diversity, and clinical adoption, concluding with perspectives on future research. Recent years have seen major strides in the automated analysis of thyroid nodules using Convolutional Neural Networks (CNNs) applied to ultrasound and cytological images. While earlier computer-aided systems relied on handcrafted features and were limited by operator variability, CNN-based deep learning provides both autonomous feature extraction and improved diagnostic accuracy. This survey paper systematically reviews the evolution of automated thyroid nodule analysis, highlights the latest technological advancements including multimodal models and explainable AI and identifies key challenges and future directions for clinical translation.
2. Background CNNs are a class of deep learning models particularly suited for image-based tasks owing to their automatic feature extraction capabilities. Various studies have shown CNNbased computer-aided diagnosis systems can match or even surpass expert radiologists in the accuracy of thyroid nodule classification on ultrasound images.
2.1 Evolution of Automated Thyroid Nodule Analysis 2.1.1 Early Computer-Aided Systems Initial research into AI-assisted thyroid diagnosis in the 1990s focused on manually engineered features (e.g., nodule borders, echogenicity) analysed by shallow machine learning models such as support vector machines and early neural networks. These systems improved diagnostic consistency but suffered from limited specificity, operational complexity, and dependence on subjective feature extraction.
2.1.2 Shift to Deep Learning and CNNs
Key Words: Thyroid nodule, Deep learning, Convolutional Neural Networks, Ultrasound, ComputerAided Diagnosis, Artificial Intelligence, Cytology, MultiModal AI
Deep learning, particularly Convolutional Neural Networks, marked a paradigm shift: they automatically learn hierarchical features from raw image data, enabling more objective, accurate, and scalable analysis.
1. INTRODUCTION
CNN models like VGG-16, ResNet, and custom medical architectures have demonstrated high performance in thyroid nodule detection, classification, and risk stratification on ultrasound and cytological images.
Thyroid nodules are extremely common, and accurate differential diagnosis is crucial to avoid unnecessary surgeries while promptly identifying malignancies. Traditional diagnosis relies upon subjective radiological and cytological assessments, contributing to inter-operator inconsistency and over-diagnosis.
2.2 Advancements and Prospective Directions 2.2.1
AI-driven computer-aided diagnosis, particularly deep CNNs, have emerged as robust solutions, outperforming or matching the accuracy of expert radiologists by automatically extracting discriminative features from medical images. Recent advances in deep learning, particularly CNNs, have demonstrated high accuracy in extracting complex features
© 2025, IRJET
|
Impact Factor value: 8.315
|
Multimodal and Multitask CNNs
Integration of multimodal data-including genetic, clinical, and imaging information— yields superior diagnostic performance and lays the foundation for precision medicine in thyroid care.
Multi-task learning frameworks can simultaneously perform detection, classification, and even mutation prediction, leveraging shared representations.
ISO 9001:2008 Certified Journal
|
Page 215